There was a good epidemiological study on stress in MS. It had a good method, almost. People in a registry, who had experienced the death of a child, were examined to see if they had any more MS than the rest of the population. Yes, they did have slightly more.
The thing is, MS causes people tp have lousy cognition. MS ranges from severe (lethal) to being only occasionally symptomatic, and very benign. MS is a little bit heritable. What if the kids of these people, rarely, had a subtle form of undiagnosed MS, and this caused them to die in accidents, such as pedestrian-vs-car accidents? Accidents are a serious cause of death in non-elderly people: 40,000 people die per annum in the US due to car accidents, and this has been so since the 1950s (total of two million deaths).
I would like to look at that study's data to see about the matter of accident deaths. And there are other possible confounds that might mess it up, which would take forever to discuss.
This is pretty much the best and most objective kind of psychopathological study I know of -- and i have reservations about it still. I dont think there many more types of really objective studies that can be done in that field. I havent seen any. When you ask the sick person, after getting sick, what they remember, they are very biased. The typical neurologically ill person is hypersensitive, not only to what is around him, but to his past. And there is a burning drive, part of human nature, to discover the cause of the waylaying illness. (I myself was at one time very vulnerable to this.) Thats why so many people are convinced that they were sickened by X, Y, or Z, when this really cant be proven from a single case, unless they got sick the very instant that X Y or Z occured. Otherwise it is really hard to rule out coincidence. There are over a billion english speakers in the world, and of course there are 6 billion people overall. That provides a lot of opportunity for coincidence.